WD Red Drives might have similar problem as WD Green

[anRossi]

I bought 4 WD Red 1TB drives about a month ago and built a FreeNAS box around them.

After burning them in, they had a load cycle count (S.M.A.R.T. attribute 193) of about 100 (101, 105, 106, 90).

It's been a few weeks and the box has spent most of that time off. I turn it on for a day every few days and back up to it, then I shut it off. The backups happen twice a day, 12 hours apart.

I ran a smart command to check the parameters on the drives out of curiosity. In 2 week's time, the drives have jumped to almost 700! Clearly these drives are parking the heads far more often than I'd like them to.

I searched around, and found a tool WD makes specifically for the WD Red drives that is supposed to lengthen the wait time before the heads park. Annoyingly, the tool doesn't tell you what the current timeout is, or what it is lengthened to. Nor does it let you specify a custom timeout.

I'm posting it here in case anybody else has WD Reds and is worried about the drives wearing themselves out.
http://support.wd.com/product/download.asp?groupid=619&sid=201&lang=en

 

[cyberjock]

We know.. there's already 4+ threads on the topic as well as a thorough how-to guide written by me on this topic on how to fix the problem. If you check out my how-to you can figure out for yourself what the timeout is.

 

[anRossi]

Oops. Guess I got a little excited about my discovery and didn't think to search before posting.
Thinking about it, I vaguely recall reading some mention of this; I was wondering if I had read about it here or somewhere else. Honestly, there's been so much forum reading that I can't keep straight where I've read anything.
I'll look for your how-to-guide. It uses wdidle3 or idletools, right?

 

[cyberjock]

wdidle3.

I've always questioned the actual differences between Reds and Greens(and if those differences constituted the price difference). It started with appearing to be different hardware, but now it's looking more and more like Reds and Greens are the same hardware, and they are exceedingly close in the firmware too!

Nobody has really provided objective quality evidence that they aren't the same today, and while I can't validate they are the same, we can't validate that they aren't the same either. So if you feel the Reds give you a good bang-for-your-buck then go for it!

 

[anRossi]

I found your post (here, for posterity http://forums.freenas.org/index.php?threads/hacking-wd-greens-and-reds-with-wdidle3-exe.18171/ ), and I hadn't seen it before, but maybe I had read reference to it. Reading now.

I had been doing drive research in January, went to my local PC store for other parts and they were having a sale on the WD Reds. I figured, these were drives I was considering anyway, I didn't have to worry about shipping as much (since these were retail boxes, they probably arrived as a pallet and not individually kicked by shippers), and the price was right, so I bit.

One of the drives definitely was manufactured differently than the other three; it looks like they skimped on it. One of them performs worse (more slowly) than the rest of them. I haven't proven if it's the "different" one or not. Two of them behave nearly identically, and one of them is notably better than the other 3. The serial numbers also tell a similar story: two are very close to each other, one is a few thousand earlier than the close pair, and one is a few thousand later than the close pair.
The best and worst are within 7% of each other in dd read/write tests, so I guess that's acceptable.

I don't really care if they're just relabeled greens or not. The warranty is worth it, and if they actually have extra vibration tolerance, that's good too. Knowing what I know now, I might have spent more and bought hitachi, and bigger capacities.
My goals at the time were a "small, simple drive" (single platter, basically). These fit the bill.

 

[panz]

wdidle3.

I've always questioned the actual differences between Reds and Greens(and if those differences constituted the price difference). It started with appearing to be different hardware, but now it's looking more and more like Reds and Greens are the same hardware, and they are exceedingly close in the firmware too!

Nobody has really provided objective quality evidence that they aren't the same today, and while I can't validate they are the same, we can't validate that they aren't the same either. So if you feel the Reds give you a good bang-for-your-buck then go for it!

I'm confused now :) because – reading the Backblaze blog – I thought that they were using both Red(s) and Green(s) and only Greens had problems:

"The drives that just don’t work in our environment are Western Digital Green 3TB drives and Seagate LP (low power) 2TB drives."

http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/

 

[cyberjock]

That's what it sounds like. But who knows? I have 24 drives in my system and I've had amazing reliability from them compared to the average user. I also make what I consider smart conservative decisions with regards to testing and taking care of the drives too.

 

[panz]

I chose all the hardware for my build (Supermicro X9SCM-F; Xeon 1230V2; 32 GB of Kingston RAM; Gooxi RM4024-660-BX case; Corsair HX650 Gold PSU).

Now I'm deciding about the drives: the more I read, the more I'm confused (last month I changed my mind about Seagate drives after reading a bad experience of yours).

Now Amazon offers 4TB Reds for € 150, which here is a pretty good price.

 

[cyberjock]

I'd have no reservations about buying WD Reds or WD Greens. Basically the day I go to buy them it'll be based on what I can firmly prove makes one better than the other. If the Reds are faster and it's clear they are faster, then I might buy them. Otherwise I'll be going with Greens.

 

[crumbz]

I just rebuilt my FreeNAS server and went with 8 x 4TB RED drives. Nothing so far has made me regret that decision.

 

[anRossi]

Hmm, in the last 30 power-on hours, my four drives have each had a maximum of 27 (and a minimum of 24) load cycles.
In this period of time, there have been 2 backups to the NAS, 12 hours apart.
Is that fairly normal? Or is that high?

 

[crumbz]

I don't know what's normal, but for comparison my drives LCC increases by 4-5 for every Power On Hour.

 

[cyberjock]

There is no "normal". It's totally dependent on your server usage.

But, at 30 a day, do you think you'll hit 250k in the next 5 years? ;)